Both heavy chains of the IgG molecule are of the γ-class and are identical to each other. The two light chains are also identical to each other—either κ or λ. Each heavy chain has a molecular weight of approximately 55,000 and consists of 440–450 amino acids. Each has a variable (V_H) region in which many amino acid substitutions provide each γ chain with binding specificity, and a constant (C_H) region which defines the specific class and subclass and imparts the biologic activity of the antibody molecule. The variable region of the γ-chain consists of approximately 110 amino acid residues, beginning at the amino (NH₂) terminal of the chain. The constant region of the γ-chain consists of 310–330 amino acids and contains three homology regions or domains, designated Cγ1, Cγ2, and Cγ3. Homology region Cγ1 is in the Fab fragment (Fig. 29.1) and extends from about amino acid 110 to amino acid 220; Cγ2 and Cγ3 are in the Fc fragment, Cγ2 extending from approximately amino acid 220 to amino acid 330 and Cγ3 from there to the COOH terminus.

Immunological analysis of myeloma proteins has disclosed four distinct subclasses of γ-heavy chains, designated IgG1, IgG2, IgG3, and IgG4 (Table 29.3). The average biological half-life of IgG1, IgG2, and IgG4 myeloma proteins injected into patients with malignancies is 21 days, whereas IgG3 has a half-life of 7–8 days. This shorter time is probably due to its fourfold longer hinge region, which appears to be easily accessible to proteolytic enzymes. In addition, IgG3 contains a large number of disulfide bonds. A wide angle of 45°–180° in IgG3 indicates a mobile molecule. Although the hinge region in the IgG3 molecule is four times longer than that in IgG1, it exists in a compact state. In the extended confirmation, the hinge region is 7–9 nm and the length of the intact molecule is only 25 nm [7]. IgG3 does not bind staphylococcal protein A except in rare cases [8]. IgG1 and IgG3 fix complement readily via the classic pathway, but IgG2 does so less well.

Although many antigens evoke an antibody response within the IgG subclasses proportional to their distribution in normal serum, others produce antibodies mainly within one subclass. For example, almost all antibodies to factor VIII (found in hemophilic persons, postpartum women, and elderly patients) are of the IgG4 subclass, while without predilection only 3–4 % of the antibodies would be of this subclass [9]. Most antibodies to dextran, levan, and teichoic acid are of the IgG2 subclass. Antibodies developed against polysaccharide antigens are also primarily found in IgG2 [10].

Although it has been reported that the serum antiplatelet antibody was restricted to the IgG3 subclass in idiopathic thrombocytopenic purpura (ITP), all four IgG subclasses were bound to platelets of ITP patients in the same distribution as in normal serum [11]. Presumably, these differences represent nonspecific binding to platelets.

Although specific IgG subclass deficiencies have been associated with increased infections, the clinical impact of selective subclass deficiencies is unclear. Among a group of patients with selective IgA deficiency, those with reduction of IgG2 had frequent respiratory infections, but those with normal IgG2 levels did not [12]. In 139 patients with isolated IgA deficiency and an increased susceptibility to infection, IgG4 deficiency was more common than in controls. Contrary to the prior study, there was no significant association between IgG subclass deficiency and increased susceptibility to infection [13]. The IgG2 subclass is thought to be most important in protecting against bacterial sepsis [14]. In a series of 22 patients with ataxia–telangiectasia, very low or borderline-low serum IgG2 levels were found in all, and IgG4 was undetectable in 19 of the 22 patients. Nine of the patients with an undetectable or very low IgG2 level had chronic lung disease or chronic sinusitis, or both [15]. In another series, 4 of 422 unselected patients were found to have an isolated absence of IgG4, as determined by a ­sensitive radioimmunoassay. All had severe recurrent sinopulmonary infections [16]. Others have suggested that deficiency of IgG3 is becoming more common and is more frequently associated with mild infections [17]. In a group of 89 patients with bronchiectasis, the IgG1, IgG2, and IgG3 levels were increased compared with controls. There was an overall incidence of deficiency of 1 % for subclasses IgG1, IgG2, and IgG3 and 5 % for IgG4. The authors concluded that IgG subclass deficiency was comparatively rare in bronchiectasis [18]. Extensive reviews of IgG subclass deficiency have been published [19–21].

We have found that the distribution of IgG subclasses did not differ among patients with multiple myeloma and those with monoclonal gammopathy of undetermined significant (MGUS; benign monoclonal gammopathy) [22]. There was also no relationship between clinical and laboratory features in multiple myeloma of different subclass types. However, patients with a monoclonal protein of different IgG subclasses have greatly different patterns of suppression. For example, patients with IgG2 monoclonal proteins are more likely to have repressed residual IgG than are patients with IgG3, IgG1, or IgG4 [23].

Radial immunodiffusion in agar with appropriate antisera can be used for semiquantitative concentrations of IgG subclasses in sera, but nephelometry is now commonly used for quantitation of IgG subclasses. These procedures are similar to those used for IgG, IgA, and IgM assays, and they are rapid and reliable.

Historically, the light chains were encountered in 1845, when Macintyre [24] and Henry Bence Jones [25] noted that the urine of a patient with multiple myeloma precipitated when heated, cleared when boiled, and reprecipitated when cooled. In 1962, Edelman and Gally [26] demonstrated that the light chains prepared from a serum IgG myeloma protein and the Bence Jones protein from the same patient’s urine not only had the same thermal solubility but also had similar spectrofluorometric behavior, the same molecular weight, identical appearance on chromatography with carboxymethylcellulose and on starch gel electrophoresis after reduction and alkylation, the same ultracentrifugal pattern, and the same amino acid sequence. Thus, it was found that the urinary proteins that evoked such interest from Macintyre and Bence Jones were the light chain components of the immunoglobulin molecule.

Two distinct groups of Bence Jones proteins (group I and group II) were recognized by Bayne-Jones and Wilson [27] in 1922. In 1956, two major classes of Bence Jones proteins were described by Korngold and Lipari [28], and these have been designated κ and λ. Approximately two-thirds of serum IgG monoclonal proteins are of the κ-type and the remainder are λ. In patients with Waldenström’s macroglobulinemia, approximately 80 % are κ, but the opposite occurs in IgD myeloma, in which 60 % are λ [29].

Immunoglobulin light chains have a molecular weight of 22,500 and contain 210–220 amino acids. Analysis of amino acid sequence in individual light chains has disclosed a constant and a variable domain. The region of the chain from amino acid 107 to the COOH terminus at position 210–220 is very similar in chains of the same type (κ or λ) and has been designated as the constant domain (C_L)—although amino acid substitutions have been found at a dozen different positions in the constant region of human λ chains [30]. The region from the NH₂ terminus (position 1) to approximately position 107 has been different in every light chain thus far analyzed, and it is called the variable domain (V_L). The variable half of the light chain contains the unique thermal solubility characteristics of Bence Jones proteins [31].

Amino acid sequence studies have revealed four basic groups of κ-light chains, designated κI, κII, κIII, and κIV [32]. Their frequencies of occurrence are approximately 60 %, 10 %, 28 %, and 2 %, respectively, and as expected, these figures reflect the proportion of κI, κII, κIII, and κIV light chains in immunoglobulins of normal serum [33]. In a series of 23 patients with acquired Fanconi syndrome, 96 % had a monoclonal κ-protein [34]. In another series, eight of nine patients with acquired Fanconi syndrome had κ1-subtype [35]. In light chain deposition disease, the light chain is κ in 90 % of patients, but the κIV subclass is overrepresented [36, 37]. Six subclasses of λ have been reported with the following distribution: λI, 26 %; λII/λV, 38 %; λIII, 22 %; λIV, 3 %; and λVI, 11 % [38]. There is a preferential association of λVI subclass with primary systemic amyloidosis (AL) [39]. In fact, with rare exceptions, all patients with monoclonal λVI subclass have amyloidosis.

Ordinarily free κ-chains (unbound to heavy chains) are in the monomeric form (molecular weight 22,500), but they may exist as noncovalently linked dimers or as a mixture of monomers and dimers, whereas free λ-chains occur as dimers (molecular weight 45,000) linked covalently via their penultimate cysteinyl residues [33]. In addition, κ-light chains precipitate maximally over a narrower range of pH than do λ-light chains [40]. Recognition of light chains as κ or λ by monospecific antisera depends on the variable and constant domains as well as the interdomain “switch” region between the variable and the constant regions [41].

Bence Jones proteins are not degradation products of the complete immunoglobulin in the serum but are synthesized by plasma cells in molar excess compared to heavy chain synthesis. Putnam and Hardy [42] showed that ¹³  C-labeled glycine is incorporated promptly into Bence Jones protein and cannot be the result of breakdown of serum or other tissue protein. Light chains clearly are catabolized by the kidney. After injection into mice, ¹³¹I-labeled Bence Jones proteins of humans were cleared rapidly from the circulation. Ligating and severing the ureters lengthened the disappearance time only slightly, but removal of the kidneys greatly prolonged the survival of light chains, a finding indicating the major role of renal catabolism. Human IgG was not cleared by the kidney in the same experiments [43].

IgG catabolism is increased at high protein concentrations. At low IgG concentrations, most of the IgG is phagocytized into vesicles, where it is protected from catabolism. When the IgG concentration is high, significant amounts of IgG molecules will not be bound by recycling receptors and will be degraded, the result of which is a high catabolic rate [44–46].

Treatment of IgG with the proteolytic enzyme papain cleaves it into three pieces: two Fab fragments, so named because they possess antigen-binding activity, and an Fc fragment, because in certain species it can be crystallized (Fig. 29.1). Each Fab fragment (molecular weight 52,000) consists of the amino-terminal region of a γ heavy chain and a complete light chain. The hinge region, consisting of 15 amino acid residues between Cγ1 and Cγ2, is a flexible junction between Fab and Fc fragments. This hinge permits the angle between the Fab fragments to vary from a few degrees to almost 180°.

Three regions of “hypervariability” have been identified in V_L: positions 24–34, 50–56, and 89–97 [47]. This hypervariability and the existence of similar hypervariable regions in the variable region of the heavy chain (V_H) in positions 30–37, 51–68, 84–91, and 101–110 permit the formation of many different antigen-combining sites—each consisting of approximately 30 amino acid residues [48]. These hypervariable regions constitute the antigen-binding site, and the three segments are called complementarity-determining regions, or CDRs (CDR1, CDR2, and CDR3).

The Fc fragment (molecular weigh 48,000) consists of the C-terminal region of both heavy chains, which are linked to each other by disulfide bonds. Biological activities of the Fc fragment include fixation of complement; passive cutaneous anaphylaxis (PCA); binding with rheumatoid factors; binding with receptors on macrophages, monocytes, neutrophils, and lymphocytes; reaction with staphylococcal A protein; and transfer across the placenta. Isotypic specificity—specificity for γ, α, μ, δ, or ε—resides in the Fc fragment. The Fc piece is devoid of antigen-binding activity.

Secretory IgA (SIgA) (Fig. 29.2) is found in high concentration in the secretions of many glands lining the respiratory and gastrointestinal tracts and in tears, colostrum, and urine. SIgA has a sedimentation coefficient of 11S and a molecular weight of 390,000. It is composed of two IgA molecules attached by disulfide bonds to a glycoprotein (molecular weight 60,000–80,000) called the secretory component. Secretory component increases the resistance of the molecules against digestion by trypsin and pepsin. The secretory component is synthesized in mucosal epithelial cells. In one unusual case, cancer cells from bronchogenic adenocarcinoma produced large amounts of secretory component, resulting in high serum levels of SIgA [60]. Serum IgA does not make a significant contribution to IgA in gastrointestinal fluids [61]. In both the gastrointestinal and respiratory tracts, dimeric IgA with J chain (described in the next section) is synthesized in plasma cells and secreted, after which it combines with secretory component on the epithelial cell surface. SIgA is then transported through the cytoplasm in endocytic vesicles to the luminal surface [62].

SIgA exhibits antibacterial and antiviral activity, and evidence indicates the SIgA antibodies are stimulated preferentially by upper respiratory tract infections. SIgA has a neutralizing effect on viral replication [63].

In secretions, IgA prevents microorganisms and foreign proteins from adhering to and penetrating the mucosal surfaces [61]. SIgA provides the most potent immunoglobulin signal for the degranulation of eosinophils. It may be the principal immunoglobulin that mediates eosinophil effector function at mucosal surfaces [64].

In contrast to that in serum IgA, approximately 50 % of SIgA is of the IgA2 subclass. Approximately 95 % of IgA in secretions is polymeric, whereas 90 % of serum IgA is monomeric. A patient with no secretory component in the saliva who presented with chronic intestinal candidiasis has been described [65].

The intestinal lamina propria cells produce the largest amount of polymeric IgA [66]. In a patient with pronounced serum IgA deficiency, most of the serum IgA was monomeric and from tonsillar IgA-producing cells [67].

The joining (J) chain is a non-immunoglobulin protein with a molecular weight of approximately 15,000 [68]. It is bound by disulfide bridges to the Fc portion of the heavy chain and thus forms the dimers or polymers of IgA and the pentamer structure of IgM. It is never found in IgG or in monomers of IgA proteins, and yet J chain is found in cells secreting IgG. Only one J chain is present in each polymer of IgA or pentamer of IgM. The J chain of IgM is identical to that of IgA. The lack of J chain reduces the ability of IgA1 to bind secretory component and interferes with the epithelial transport of polymeric IgA [69].

Although monoclonal proteins had long been considered abnormal, studies have indicated that they are excessive quantities of immunoglobulins that occur normally. The striking feature of monoclonal proteins that led investigators to consider them abnormal is their homogeneity. Whereas the population of normal IgG in human serum has diverse binding activity and is electrophoretically heterogeneous, IgG monoclonal proteins are localized sharply in their electrophoretic migration.

Kunkel [125] showed that each heavy chain subclass and light chain type in monoclonal proteins has its counterpart among normal immunoglobulins and also among antibodies. After the discovery of the two types of light chains (κ and λ) in myeloma proteins in a ratio of approximately 2:1, these same light chains were detected in essentially the same ratio among normal immunoglobulins. Similarly, the IgG and IgA subclasses and IgD class were discovered among myeloma proteins and then found as normal serum components. These identifications reversed the belief that the monoclonal proteins of multiple myeloma and macroglobulinemia are abnormal and have indicated that they represent the overproduction of a normal product by a clone of abnormal proliferating cells. However, many of the heavy chains found in the heavy chain diseases show significant deletions of amino acids: they are abnormal immunoglobulins.

Even the antigenic determinants (“idiotypic specificities” or “individual antigenic specificities”) that are associated with the binding sites were believed to be associated uniquely with myeloma proteins but have been shown to occur among antibodies [126]. Conversely, studies of highly purified antibodies have revealed a homogeneity approaching that seen in monoclonal proteins. In some instances, after primary immunization with a carbohydrate antigen, monoclonal immunoglobulins consisting of a single light chain type and single IgG heavy chain subclass have been seen [127].

The possibility expressed by Kunkel [125] that myeloma proteins are individual antibodies and are products of the individual plasma cells arising from a single clone of malignant cells has been supported by their antigen-combining activity [128]. Monoclonal antibody activity in humans has been associated with cold agglutinin disease as well as with a wide variety of bacterial antigens including streptolysin O, staphylococcal protein, Klebsiella polysaccharides, and Brucella [129]. It is almost certain that more human monoclonal proteins will be found to have antibody activity. Indeed, it has been postulated that all myeloma proteins may have such activity [130].

In this view, as illustrated in Fig. 29.7, the normal assortment of IgG molecules comprises minute amounts of homogeneous proteins from many diverse single clones of plasma cells and so is called polyclonal. If a single clone escapes the normal control mechanism, excessive proliferation occurs, leading to synthesis of excess protein with a single heavy chain class and subclass and light chain type that is now distinct from the normal immunoglobulins. More than 60 years ago, Gutman [131] used the term M component (or peak) in describing a monoclonal protein. This monoclonal protein is associated with abnormal plasma cell proliferation and may be associated with a neoplastic process. Support for the one-cell, one-immunoglobulin concept comes from the studies of the cellular localization of these proteins. Experiments performed with antisera to light chains have shown that nearly all individual plasma cells contain κ- or λ-light chains but not both.

Normally plasma cells produce heavy chains and a slight excess of light chains that spill over into the urine and are catabolized. In IgG myeloma, greater than three-fourths of patients have an excess of light chains that may be excreted into the urine (Bence Jones proteinuria) or catabolized [132]. A small excess of light chain production could be due to imbalance in translation or transcription within the clone or a possible suppression of heavy chain synthesis in the clone. In other instances, heavy chain is not produced by the plasma cell and only excessive quantities of light chains are detected (“light chain multiple myeloma”) [133]. Finally, a small proportion of myeloma cells do not secrete either heavy or light chains in detectable amounts, whether because of a simple failure of synthesis or a blocking of secretion, and these are called nonsecretory. In heavy chain disease, portions of the heavy chains of IgG, IgA, or IgM are present in serum or urine.

Analysis of the serum or urine for monoclonal proteins should be performed with a sensitive, rapid, and dependable screening method to detect and quantitate the presence of a monoclonal protein and a specific assay to identify it according to its heavy chain class and light chain type [134–137]. Electrophoresis with high-resolution agarose gel is preferred for screening and should be done in conjunction with serum FLC quantification [138, 139]. After electrophoresis, immunofixation should be used to confirm the presence of a monoclonal protein and to distinguish the immunoglobulin heavy chain class and its light chain type. Due to their higher sensitivity for obscured abnormalities and monoclonal FLC, immunofixation and quantitative FLC also should be done if multiple myeloma, macroglobulinemia, primary amyloidosis, or a related disorder is suspected [137–139], even if the agarose gel electrophoresis does not show a localized band or spike.